This release contains many updates to the functionality of the package. The highlights:
- New processes for pre-checking variables included, including standardization of variables.
- Ability to specify keyword arguments for the underlying scikit-learn logistic regression that is implemented under the hood when doing testing of variables.
- An updated matching algorithm that utilizes binary search to deliver matching in O(logN) time. This brings match times down from ~200ms a match (on my local machine) to 20 nanoseconds.
- The ability to specify a caliper in the matching process such that matches outside of a window are thrown away.
- Improved documentation.